Saved in:
Bibliographic Details
Main Authors: Yu, Tianyu, Zhang, Haoye, Li, Qiming, Xu, Qixin, Yao, Yuan, Chen, Da, Lu, Xiaoman, Cui, Ganqu, Dang, Yunkai, He, Taiwen, Feng, Xiaocheng, Song, Jun, Zheng, Bo, Liu, Zhiyuan, Chua, Tat-Seng, Sun, Maosong
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2405.17220
Tags: Add Tag
No Tags, Be the first to tag this record!
Table of Contents:
  • Traditional feedback learning for hallucination reduction relies on labor-intensive manual labeling or expensive proprietary models. This leaves the community without foundational knowledge about how to build high-quality feedback with open-source MLLMs. In this work, we introduce RLAIF-V, a novel framework that aligns MLLMs in a fully open-source paradigm. RLAIF-V maximally explores open-source MLLMs from two perspectives, including high-quality feedback data generation for preference learning and self-feedback guidance for inference-time scaling. Extensive experiments on six benchmarks in both automatic and human evaluation show that RLAIF-V substantially enhances the trustworthiness of models at both preference learning and inference time. RLAIF-V 7B reduces object hallucination by 80.7\% and overall hallucination by 33.7\%. Remarkably, RLAIF-V 12B further reveals the self-alignment potential of open-source MLLMs, where the model can learn from feedback of itself to achieve super GPT-4V trustworthiness.